helio-mcp
Pass
Audited by Gen Agent Trust Hub on Aug 3, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill facilitates the ingestion of external research data, creating a surface for indirect prompt injection where adversarial content in reports or responses could influence agent behavior. \n
- Ingestion points: Data is retrieved through tools such as
get_test_report,get_filtered_responses, andget_responsesinreference.md. \n - Boundary markers: No specific delimiters or instructions are provided to help the agent isolate external data from its core logic. \n
- Capability inventory: The agent has access to state-changing tools like
create_test,update_test, andsend_test. \n - Sanitization: The instructions do not define methods for validating or filtering data retrieved from the Helio platform. \n- [EXTERNAL_DOWNLOADS]: The documentation describes fetching the
@zurb/helio-mcppackage from the npm registry for installation. \n- [REMOTE_CODE_EXECUTION]: The configuration examples utilize thenpxutility to download and run the vendor's server code at runtime. \n- [COMMAND_EXECUTION]: The skill instructions involve configuring shell commands, such asnpx, to run the MCP server.
Audit Metadata